We propose to conduct the first large scale study of risk factors for human immunodeficiency virus (HIV)-related non-Hodgkin's lymphoma (NHL) in children. This molecular epidemiologic investigation will employ a nested case-control design to determine the relationship between the risk of developing NHL and the host viral burden at the time of lymphoma diagnosis. This study will include the majority of newly diagnosed children with HIV- related NHL in the U.S. and Canada and serve as a companion to the NCI and NIAID co-funded Pediatric AIDS/Lymphoma Network grant. We hypothesize that HIV-induced immunodeficiency combined with a high viral burden from HIV, EBV, CMV and HHV6, or EBV-induced chronic B-lymphocyte stimulation predispose to the development of NHL. In addition, Mycoplasma may act as a co-factor with HIV in increasing the risk of NHL. Cases will be newly diagnosed patients with NHL confirmed by centralized pathology review with a documented HIV infection, not previously treated for lymphoma. We plan to accrue 80 cases <15 years of age restricted to perinatal, blood or blood product transmission. Two groups of controls will be selected. For the first, 320 matched controls will be selected from the cohort of HIV-infected patients followed at participating centers with no evidence of malignancy. Controls will be matched to cases by source of HIV infection and age. An additional control group consisting of 160 children with NHL without HIV infection (HIV-NHL+) will be matched to cases (HIV+NHL+) by NHL histology, immunophenotype and age. This alternate control group will be compared to the case group to determine if the pathogenesis of NHL in HIV+ children is similar to that for HIV- children. Specimens from cases (HIV+NHL+) and controls (HIV+NHL-) will be assessed in a central laboratory for evidence of viral and Mycoplasmal infection and immune competence as well as other potential risk factors, previous anti- retroviral therapy, anti-infection therapy and opportunistic infections. Using quantitative PCR and serology, subjects will be compared with respect to the presence and quantity of co-infection at the time of lymphoma diagnosis with Mycoplasma, EBV, CMV and HHV6. Tumor tissue for cases and the second control group (HIV-NHL+) will be compared to characterize immunoglobulin gene rearrangements and c-myc oncogene translocations to determine if monoclonal or polyclonal expansion has occurred. If EBV is present, evidence for clonal proliferation and transforming gene activity will be determined. Rearrangements and enhanced transcription of c-myc have been demonstrated in adult patients with AIDS-related lymphomas. It will be important to further elucidate pathways for tumor formation in HIV- positive children. This study will provide a unique opportunity to study HIV-related lymphomas in children just as the malignancy rate is starting to rise.